How to deal with NA in correlations CIs in Saturated model

Hi OpenMx community,

I am new to using OpenMX. I am trying to run a saturated trivariate model on my twin data. The estimation of the model and its submodels with constraints works well; however, when I attempt to obtain the correlation indices derived from the final model with all constraints, most of the confidence intervals return as NA.

In a previous study using the same sample and scales very similar to those used in this study (only with a slightly higher number of items), this issue did not occur. What mistake might I be making?

I should mention that I am using the latest version of OpenMX (v2.21.13) with the SLSQP estimator. I tried switching to NPSOL following the procedure recommended by you, but I encountered the following error:

Error in file(filename, "r", encoding = encoding): cannot open the connection to 'https://vipbg.vcu.edu/vipbg/OpenMx2/software/getOpenMx.R'.

Here is the script I've been using